Online retailers often use a recommendation system to suggest new products to consumers. Consumers are compared to others with similar characteristics such as past purchases, age, income, and education level. A data set, such as the one shown in the accompanying table, is often used as part of a product recommendation system in the retail industry. The variables used in the system include whether or not the consumer eventually purchases the suggested item (Purchase = 1 if purchased, 0 otherwise), the consumer’s age (Age in years), income (Income, in $1,000s), and number of similar items previously purchased (PastPurchase). 

a. Perform KNN analysis on the Retail_Data worksheet to determine whether or not a consumer is likely to make a purchase. Score the records of the 10 new consumers in the Retail_Score worksheet. What is the optimal k? What is the predicted outcome for the first new consumer? 

b. What is the misclassification rate for the optimal k? 

c. Report the accuracy, specificity, sensitivity, and precision rates for the test data set (for Analytic Solver) or validation data set (for R). 

d. Generate the decile-wise lift chart. What is the lift value of the leftmost bar? 

e. Generate the ROC curve. What is the area under the ROC curve (or the AUC value)? 

f. Comment on the performance of the KNN classification model. Is the KNN classification an effective way to develop a recommendation system?

Transcribed Image Text:

Purchase Age 48 47 1 1 ⠀ 0 ⠀ 34 Income 99 32 1 110 PastPurchase 21 0 *** 2